cvat-opencv
cvat-opencv copied to clipboard
Export annotated dataset feature not working as expected
My actions before raising this issue
- [X] Read/searched the docs
- [X] Searched past issues
I am having trouble in exporting the annotated dataset, when trying Segmentation Mask 1.1 format. I keep getting only 3 items in the zip - labelmap.txt, JPEGImages, ImageSets.
I dont see SegmentationClass
or SegmentationObject
items in the zip at all, despite my having labeled a few objects in the images. This is a consistent behaviour. Anything I might be missing?
Expected Behaviour
The exported zip should contain the 5 items - labelmap.txt, JPEGImages, ImageSets
, SegmentationClass
and SegmentationObject
Current Behaviour
I keep getting only 3 items in the zip - labelmap.txt, JPEGImages, ImageSets.
Steps to Reproduce (for bugs)
- Create a new labeling job by uploading a sequence of images.
- Draw polyline around the objects in the image. (Shape, not Track)
- Save
- Export Task dataset > Format : Segmentation Mask 1.1 (Include images or not) > enter zip name > OK
Context
I am not able to export the labeled dataset, and do not know where to look for logs for this process specifically. docker logs
doesnt give any helpful logs unfortunately.
Your Environment
- Git hash commit (
git log -1
): d7560bbd39fec68f944515c2591dda74f3764b90 - Docker version
docker version
(e.g. Docker 17.0.05): 20.10.17 - Are you using Docker Swarm or Kubernetes? No
- Operating System and version (e.g. Linux, Windows, MacOS): Linux
- Code example or link to GitHub repo or gist to reproduce problem: N/A
- Other diagnostic information / logs:
Logs from `cvat` container
@sebkken
Draw polyline
A polyline is infinitely thin. It will not be drawn on a segmentation mask. Use polygons instead.
@sebkken, could you describe what is the expected result? Currently, we only export 2d annotations with non-zero area, i.e. boxes, polygons and ovals (docs).
@bsekachev @zhiltsov-max Thanks for the information. We started using the polygons instead of polylines, and this issue has been resolved. Thanks so much for the prompt response.